Hello all, It's a very difficult time now, as the covid-19 cases are going up like anything here in India and the world, we have lost many dear ones due to this deadly disease. It's been very difficult to keep mind and body to do the required work. Also due to continuous lockdown, it is very difficult to source the parts. In this project, I'm trying to work on my POC with available resources. Please excuse me for not using standard parts. Please feel free to read and suggest the enhancements.
Initial Pitch :
Agriculture is one of the noble professions by which one can feed themselves and also produce enough to feed others. At the end of the day it boils down to the food, in order to live we need food and we cannot eat money! It's very dangerous to see that trend that is been affecting the farmlands which are being converted into urban ecosystems and Special Economic Zones or SEZ. Which are one of the main reasons for the increase in commodity prices and global warming.
Even with a very less amount of farming land and very fewer farmers, we are facing issues such as pest infestation, animal infestation and over rainfall or drought, etc.
In this project, I would like to address the issues such as anima infestation and pest infestation detection using the microphone and Arduino nano IoT.
The main idea here is to detect the animal/pest infestation in order to protect the crops. With the help of TinyML and Edge Impulse,
we will use the Microphone and temperature sensor to obtain and train a machine learning model which can be run on the arduino nano 33 sense to detect the infestation of crickets and elephants.
In developing countries like India Elephants and crickets cause a lot of crop damage which can be controlled if known in advance. With the help of edge impulse trained model, we will detect the presence of these pests and elephants and then we can alert the farmers before they enter the farmland. This animal-human conflict is also responsible for the death of many humans and animals. This system is proactive rather than being reactive.
Training the voice model of elephants and crickets, File Photo: Source: Google Images
This is the basic idea behind the system.
We can use multiple parameters and sensors to predict the weather and can help the farmer to get the data on his phone via the mobile application or SMS.